File: gee_its_simple_backsolve.m

package info (click to toggle)
suitesparse 1%3A7.10.1%2Bdfsg-1
  • links: PTS, VCS
  • area: main
  • in suites: trixie
  • size: 254,920 kB
  • sloc: ansic: 1,134,743; cpp: 46,133; makefile: 4,875; fortran: 2,087; java: 1,826; sh: 996; ruby: 725; python: 495; asm: 371; sed: 166; awk: 44
file content (39 lines) | stat: -rw-r--r-- 1,345 bytes parent folder | download | duplicates (2)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
function b = gee_its_simple_backsolve (U, b)
%GEE_ITS_SIMPLE_BACKSOLVE x=U\b where U is upper triangular
% Perform back substitution to solve x=U\b.  U must be square.  The lower
% triangular part of U is ignored (this allows L and U to be packed into a
% single matrix).
%
% Example:
%
%   x = gee_its_simple_backsolve (U,b) ;
%
%   % which is the same as
%   x = triu (U) \ b ;
%
% See also: mldivide, gee_its_simple, gee_its_short, gee_its_simple_forwardsolve

% GEE, Copyright (c) 2006-2007, Timothy A Davis. All Rights Reserved.
% SPDX-License-Identifier: BSD-3-clause

%-------------------------------------------------------------------------------
% check inputs
%-------------------------------------------------------------------------------

if (nargin ~= 2 | nargout > 1)                                              %#ok
    error ('Usage: x = ge_its_simple_backsolve (U,b)') ;
end

% ensure U is square, and that U and b have the same number of rows
gee_its_simple_check (U, 'U', b) ;

%-------------------------------------------------------------------------------
% backsolve solve, overwriting b with the solution U\b
%-------------------------------------------------------------------------------

n = size (U,1) ;
for k = n:-1:1
    b (k,:) = b (k,:) / U (k,k) ;
    b (1:k-1,:) = b (1:k-1,:) - U (1:k-1,k) * b (k,:) ;
end